Industry Forum

About

IEEE ONCON 2023 will host an Industry Forum session during the conference. Industry Forum is an IES program for Industry to engage with research in a productive manner. Industry speakers are invited to discuss industry, technology directions, and, most importantly, challenges for the companies. These presentations inform the attendees on the vision and application of technologies in business and what challenges companies are encountering.  The forum also offers the opportunity for researchers to study their challenges and know the contacts in the companies should they have a solution that the company might utilize. We want all conference attendees to engage in the Industry Forum and listen to the presentations of our industry speakers so all communities can benefit. For additional Industry Forums organized in IES events visit https://www.ieee-ies.org/industry-forum


Speaker #1

Magnus Lindhé
Research Leader, Decision & Control for Cyber-Physical Systems
Ericsson Research, Stockholmm Sweden

Title: Cyber-Physical Systems Research at Ericsson

Abstract: This talk will give an overview of Ericsson’s research within cyber-physical systems. In 5G and 6G networks, machines will be a major user category making use of new capabilities such as low latency, high bandwidth, network slicing, positioning and edge compute. The cyber-physical systems team is exploring some valuable use cases for this, to get input for future standardization and build a toolbox of best practices for enterprises transitioning to 5G. I will give a few examples of use cases such as offloading of control and perception to the cloud or edge, collaborating robots, communication-aware motion planning, sensor fusion for localization and SLAM interoperability.

Biography: Magnus Lindhé is a Research Leader for the Decision & Control team at Ericsson Research in Stockholm, Sweden. He holds a MSc in Electrical Engineering and a PhD in Automatic Control with a research focus towards communication-aware robotics, both from the KTH Royal Institute of Technology in Stockholm. Between 2012 and 2022, he worked as a software architect for robotic vacuum cleaners at Electrolux in Stockholm and led the development of navigation algorithms for the Pure i9 series of robotic vacuum cleaners, of which more than 300 000 units are sold worldwide. He is a WASP industrial PhD supervisor and member of the WARA Robotics Core Team.

Speaker #2

Haifeng Wang
the chief scientist, provincial distinguished expert and talent
Fengyuan Hydrogen Energy Company, China

Title: Fuel cell systems and applications at Fenergy, a global leading supplier of hydrogen fuel cell stacks and system from China

Abstract: This talk will introduce the Fenergy way to reduce the manufacturing cost of PEM fuel cell stack so as to help speeding up the commercialization of transportation and power industry.
The Fenergy way includes focus on the research on catalyst, membrane electrodes assembly(MEA), metal bipolar plates, stacks and engine system assembly. Fenergy also strictly follow the “automotive-grade” design and manufacturing process so that the product meets the requirement of robust and durability. And finally, Fenergy creates many “commercial loops” to work closely with partners in hydrogen proudction, storage, transporation, truck&bus manufacturers and local government, so that hundreds of fuel cell products are being applied in many area in China.

Biography: Dr. Wang Haifeng is an accomplished professional with Phd, master and B.S. degrees from Tsinghua University. Additionally, he has more than 100 patents and has won the first prize in the National Final of China Innovation and Entrepreneurship Competition. He is currently the chief scientist of the key special project of the National Key Research and Development Program and holds esteemed positions such as provincial distinguished expert and talent, leader of the provincial leading innovation team, and member of various committees such as the Fuel Cell Special Committee of the China Energy Research Society, the member of the low-carbon Special Committee of the Regional Economic Society of the Chinese Academy of Social Sciences, and the member of the Tsinghua Entrepreneur Association (TEEC).

Speaker #3

Ankush Chakrabarty
Principal Research Scientist
Mitsubishi Electric Research Laboratories (MERL), USA

Title: Meta-Learning and Deep System Identification for Digital Twins

Abstract: Accurate predictive models are a core technology for any digital twin. Constructing accurate predictive models of heretofore unseen target building energy systems typically requires a large volume of dynamical data; such data is rarely available from a single source. To curtail the data required to obtain a good predictive model of a target system, we meta-learn from data archived from multiple source systems with similar dynamics.
The talk is in two parts. First, we demonstrate that meta-learning of probabilistic surrogate models improves convergence rates in Bayesian optimization, which enables few-shot calibration of building simulation models. Second, we present recent work on meta-learning deep state-space models for rapid state estimation of parameter-dependent industrial-grade vapor compression systems without explicit parameter estimation.

Biography: Ankush Chakrabarty (Senior Member, IEEE) is affiliated with the Multiphysical Systems team at MERL, with current research focusing on knowledge transfer mechanisms in machine learning to identify, optimize, and control building energy systems using limited data. Prior to MERL, he was a Postdoctoral Fellow at Harvard University, where he worked on deep learning-informed model predictive control for the development of an embedded artificial pancreas. He received his PhD in 2016 from Purdue University as a Ross Fellow in Electrical and Computer Engineering. He currently has 100+ peer-reviewed publications and 20+ US patents, and has an Erdos number of 4.


Industry forum organizers

Zhibo Pang
Sr Principal Scientist
ABB/KTH
Victor Huang
Onlye Solutions
IEEE Life Fellow
Stamatis Karnousko
SAP
IEEE Fellow
Yebin Wang
Sr Principal Research
Scientist
Mitsubishi
Meng Zhang
Professor
Xi’an Jiaotong University